The Lean Startup by Eric Ries Book Notes & Commentary
The Lean Startup Quick Summary:The Lean Startup shows how to achieve business success by learning what customers want fast. Eric Ries explains how to launch a minimum viable product, why you should ignore vanity metrics, and how to use 'split-testing' to avoid wasting years of your life building a product nobody wants.
Who is Eric Ries and why listen to him?
Eric Ries co-founded the company IMVU, a startup which lets you chat with people all over the world using your own custom 3d avatar.
Does that sound like a weird idea? Well, it was. Nobody had ever done anything like that before successfully.
That’s why the business really struggled in the beginning. They couldn’t attract any users or make a profit. But then Ries changed his approach to growing the startup and everything changed. Soon, IMVU signed up tens of millions of users and they achieved annual revenues of over $50 million. After this, Ries started advising other exciting tech startups and consulting with large companies like Intuit who wanted to know how to grow a business rapidly.
But Eric Ries was definitely NOT born a business wizard. In fact, the first two startups he was involved with failed spectacularly. It wasn’t until he found a whole new process for growing a startup that he could achieve significant success.
He began writing his lessons from starting a successful startup on a blog. This blog became very popular within the tech startup community. Eventually he compiled the best of those lessons into this book, The Lean Startup. Today, this is probably THE most recommended book for anyone wanting to launch their own startup.
Best of all, right here on this page you’ll read a fast and fun overview of The Lean Startup. I’ve pulled out some real GEMS to share with you from the book. Of course, if you enjoy this, then you’ll LOVE the book which gives many more details and examples.
Anyway, let’s jump into Lesson 1…
1. Find out what customers want as soon as possible
A startup is a human institution designed to create a new product or service under conditions of extreme uncertainty.
Most startups fail. That’s the uncomfortable truth.
This obviously goes against the romantic myth of entrepreneurship that is so common in our culture.
We’ve all heard the story before. A few college kids build the next earth-shattering company in their garage with just a couple computers and pepperoni pizzas. They persevere and become billionaires because they had just the right idea and pure passion.
Well, that’s a bunch of… you know what.
Eric Ries had the smarts, he had the ambition and he graduated from Princeton. Yet in spite of all that, his first two startup failed embarrassingly. Back then, he believed they failed because the engineering wasn’t perfect enough. But now he has knows the truth. Startups usually fail not because of poor technical execution… but because they built something nobody wanted! That is a problem all too common with startups. Teams building products through blind guesswork, hoping that customers will like it.
The solution is The Lean Startup, which is really a process to help you FIND OUT what customers really want in the least amount of time possible through small experiments.
Eric Ries says regular business plans are like launching a rocket. Everything is planned out in careful detail years in advance. However, in the unstable environment that most startups face, this kind of plan does not work at all. That’s why so many of them crash and burn. The process that Eric Ries teaches in this book is more like driving a car. Sure, you start out with a clear direction, but you can easily turn left or right at any moment as you see obstacles.
In a startup, speed is important because you only limited time until you run out of money. This is true whether you’re launching a small business with your own cash or depending on investors.
By the way, Eric Ries says a lot of this book is heavily inspired by Steve Blank. He is a serial entrepreneur and professor who created a ‘customer development process’ for startups. In fact, Steve Blank was an early investor in IMVU and he insisted their team attend his entrepreneurship course at UC Berkeley. Ries was so inspired that he carried a dog-eared copy of Steve Blank’s book The Four Steps to the Epiphany everywhere with him while building IMVU.
Most startups fail, unlike what the Hollywood movies show. Eric Ries’s first two startups failed, and he now knows it was because they built something nobody wanted! The Lean Startup is a process to avoid that happening to you, by showing a process you can follow to learn what customers really want through experiments, in the least time possible.
2. Eliminate waste and produce real value through validated learning
Lean thinking defines value as providing benefit to the customer; anything else is waste.
Lean manufacturing was a method pioneered by Toyota in the 1990s to minimize waste in their factories. This was the secret that made Toyota go from being a small company to being the largest car manufacturer in the world by 2008. Eric Ries believes that just as lean manufacturing revolutionized blue collar work, his Lean Startup methods could revolutionize white collar work, especially for new fast-growing companies.
You see, in the 1990s Japanese car makers found it difficult to compete. The big American car makers were taking advantage of mass production technique to make their cars more cheaply. But the Japanese companies like Toyota were far too small to use that method, so they decided to try something totally different which is now called lean manufacturing.
Lean manufacturing minimizes waste in factories. It helps employees waste less effort during work, and it also helps the company wastes less money by not stockpiling lots of extra inventory.
When you first hear some of the lean manufacturing methods, they sound a little crazy. For example, Toyota factories famously had an andon cord. This was basically a rope any employee in the factory could pull if they saw a defect. And when that rope was pulled, ALL production in the factory stopped so they could all focus on fixing the root cause of the defect. Most people would assume that this cord is a bad idea. I mean, if ANYBODY can pull it when there’s a little problem, then that will ruin their productivity, right?
Well, it turns out that the andon cord was a great idea! Why? Because fixing problems in the production line as they are found actually increases productivity in the long run. Every time there is a defect, Toyota wastes time or money by having to go back and fix the car or throw it out altogether. By stopping everything to fix the root causes of any defects right away, they got a lot fewer future defects. As a result, the company became a lot more productive and efficient. As a side effect, methods like this earned Toyota an enviable reputation for quality.
Waste in a startup is delivering a product or features that nobody wants.
So lean manufacturing aims to minimize waste in factories. And The Lean Startup aims to minimize waste in startups. Waste in a startup is delivering a product or features that nobody wants.
For instance, If your employee works many weeks programming some new feature, they may believe they are being very productive. BUT…. what if none of your customers actually want or will use that feature? Well, then in reality all those programming hours were wasted. In startups, this kind of “wasteful productivity” is the norm.
Eric Ries says the solution to this kind of waste is validated learning. Validated learning is a process to measure the value and quality of your product, from the perspective of your customers.
Now, the word “learning” often has a bad reputation in the the business world because learning is used as an excuse when there is lack of success. However, with validated learning you will measure the impact of every new product change with hard numbers. You’ll actually measure it with experiments, tests and metrics. This means learning is a goal we begin with at the planning stage, not just a story or excuse told after the fact.
Individual productivity is less important than learning.
When validated learning is the goal, then individual productivity of your employees is much less important. Eric Ries generally recommends startup employees work in small cross-functional teams. This is very different than the usual corporate structure that divides people into functional departments like sales, engineering, etc. However, you’ll need to set your team member’s expectations in advance by letting them know this new approach may feel less productive than before. But as Peter Drucker said, “There is nothing quite so useless, as doing with great efficiency, something that should not be done at all.”
This is really a new way of management because even tech startups are ultimately groups of humans. A new management approach is needed because startups face a situation of extreme uncertainty. In a startup, your first goal is to find out what is valued from the eyes of your customer. In other words, your first job is to figure out what your job is!
Toyota became the largest car maker in the world through lean manufacturing, a process that minimizes waste in factories. Similarly, The Lean Startup minimizes waste in startups by helping you learn what customers really want through experiments, tests and metrics.
3. Launch a minimum viable product as soon as possible
As you consider building your own minimum valuable product, let this simple rule suffice: remove any feature, process, or effort that does not contribute directly to the learning you seek.
Eric Ries says you should try to launch a basic version of your BEFORE you think you’re ready. This is called a minimum viable product (or MVP for short).
This is important because startups that wait until their product is “finished” before opening it to the public often find out they have spent months or years creating something nobody wanted. This is exactly what happened in Ries’s early startups and he was determined not to make the same mistake again with IMVU.
When IMVU launched to the public, it was full of bugs. The engineering team led by Ries actually felt embarrassed to release it! However, this gave them extra motivation to improve the product in response to customer feedback and metrics.
One bad lesson they quickly learned was they had already WASTED months working on some features that nobody wanted to use! For example, Ries himself had spent months integrating IMVU with the most popular instant messenger programs. Why? Because they assumed people wouldn’t want to sign up for a new app separately when all their friends were elsewhere. However, they saw that real users didn’t care about these integrations! They actually wanted to use IMVU to meet new friends safely, not to chat with their existing friends. This was a big shock to the team and caused them to dramatically change their plans for upcoming features.
On the other hand, their customers LOVED a stupidly simple “teleporting” feature. They actually called it better than the competition’s more technically advanced features. But the reality was the IMVU engineers had rushed this feature out at the last second and felt it was way too primitive. So they learned what customers value can be very different from what company managers, engineers or designers value.
A minimum viable product begins the learning process.
An MVP is different than a prototype because it’s actually for sale. And the moment you have a product for sale, you can begin gathering basic metrics like what percentage of your website visitors become customers. A minimum viable product can provide a much-needed ‘dose of reality’ but this is also why so many entrepreneurs hesitate to release one. They don’t want their ideas judged based on a half-finished rough product.
Early on, IMVU bought $5/day of Google ads to bring some initial visitors to their website. That was more than enough for them to get some early metrics and begin the process of optimizing their product.
When Tony Hsieh started his company Zappos, he began with a fast experiment. First he visited local retail shoe stores and asked to take photos of the shoes. He told the store owners he would put the shoes up for sale online and if anybody ordered a pair, he would come back to the store to fulfill the order. So that’s what he did. Taking photos of shoes himself and fulfilling shoe orders one-by-one at the full retail price. Obviously he didn’t make money from this, but the experiment validated his assumption that people DO want to order shoes online. This gave him confidence to then invest in a proper website, set up warehouses, hire employees, etc. Eventually Zappos became so successful that Amazon bought the company for almost $1 billion.
Launch a basic version of your product to the public, even when you don’t think it’s ready. A minimum viable product will allow you to start learning from your customer’s feedback and metrics. This is essential information to help you build the kind of product people actually want to use.
4. Evaluate each new product change with scientific split tests
When most companies release a new version of their product, they really don’t know which of the changes will be good or bad, from the eyes of customers.
For example, when Microsoft used to sell a new version of Microsoft Word annually, each new version contained dozens of new features and big changes. That was called the waterfall model of software development. The big downside of it was that customer feedback was only heard very late in the development process, when it was often too late to reverse a bad idea or poorly designed feature.
Eric Ries says a far better approach to product development is to release new features and changes individually, and split-test each one to measure the impact.
First of all, what is split testing?
Split testing began as a direct marketing technique. You see, advertisers wanted to find out what kinds of ads were most effective. So one smart copywriter had the idea to run scientific tests between different ads. So for example, they might send out two different versions of a flyer. Half of the flyers had headline A and half of the flyers had headline B, otherwise they were identical. Then they measured which headline brought more sales. How? By putting little codes in the coupons at the bottom of each flyer. When people sent in the coupons, the advertisers could then count how many sales came from each version of the flyer.
This was really a genius new development, allowing marketers to scientifically test which advertisements were most effective. This technique later became known as split-testing or A/B testing. Today with the internet, split-testing is faster and simpler than ever before.
(By the way, a great classic book that can show you how to sell with words more effectively is Scientific Advertising by Claude Hopkins. It’s a little book written decades ago, but the lessons are timeless. Claude Hopkins was one of the inventors of split-testing.)
IMVU used split-tests every day
At IMVU, Eric Ries didn’t want to spend a year working blindly on a hundred new features, only to find out half their work was unnecessary. So he tried a new product development approach based on split testing.
Changes or new features in the IMVU product were immediately released to small groups of users, picked at random. Technology made it easy to measure the impact of the new feature on their user’s actions, in terms of concrete numbers and metrics.
So they released dozens of these new changes every day, and they used technology to measure if the change had any impact on user behavior. For example, they could quickly see if a feature made users stay on their app longer or start more conversations.
This quick feedback loop helped IMVU’s engineers learn from their users directly. If people didn’t like a feature, the engineers tried to improve it or they scrapped it. If people seemed to REALLY like a feature, then the engineers planned to make more similar features in the future. This is why Eric Ries says The Lean Startup is more like driving a car than launching a rocket. With a car you can quickly change direction to a better road.
At IMVU, they split-tested every little new change and feature they made. This means they first released every new feature to a small group of random users and then measured how those user’s behavior changed. For example, did they stay longer using the app or start more conversations than usual? These metrics are invaluable for learning what your customers want.
5. Focus on actionable metrics, not misleading vanity metrics
Vanity metrics are numbers which sound impressive on the surface, but often hide the truth about a business. Vanity metrics can easily mislead and cause overconfidence.
At one point, IMVU’s vanity metrics were increasing exponentially, like their total number of users was skyrocketing. On a graph, this growth looked like a hockey stick. But Eric Ries had enough experience to know this wasn’t necessarily a great situation.
You see, IMVU was getting more users every day, but their product itself was no longer improving. Ries could only understand this when he looked at the deeper actionable metrics of the business. For example, he saw that only 1% of their free users were becoming paid users, and all of his team’s hard work over the past several months had not improved this metric even a little. It was stuck at 1%.
This situation was dangerous because without constant product improvement, their growth could easily crash. At that time, most of their users were enthusiastic early adopters, which is really a small percentage of people that would dry up eventually. If IMVU wanted to expand from early adopters to mainstream users, they would have to improve their product. And that meant focusing on the deeper actionable metrics that revealed the truth about how users interacted with their product.
Okay, I hope you’re still with me. Basically, actionable metrics are very important. Now, how can you find these actionable metrics in your own business? Through cohort analysis.
Cohort analysis is one of the best tools for measuring actionable metrics.
In a nutshell, Cohort analysis splits your users into groups/cohorts based on their actions. So it lets you to measure how different groups of users behave within your product.
For example, in IMVU Ries split his users into 5 groups/cohorts: people who registered, people who logged in at least one time, people who had one conversation, people who had five conversations and people who became paying customers.
By looking at all these cohorts of users separately, Ries could immediately see how users behaved in the product itself. More importantly, he could see whether their new product features were leading more more engagement and sales, both very important actionable metrics.
And because the actionable metrics inside IMVU were showing no improvement, the product development team became very open to user feedback. They wanted to know WHY the product was not becoming better at converting users into customers despite months of hard work. By focusing hard on these actionable metrics, the IMVU team started making real improvements to the product. Ries says this was the key step that set them on the path to making tens of millions in revenue.
Vanity metrics can sound impressive to outsiders, but they often hide the truth about a business. For example, a company can make a billion dollars in sales but still go bankrupt with no profit. On the other hand, actionable metrics measure how users behave within your product and these are critical to success. For example, measuring what percentage of free users become paid users, and working to improve this percentage.
6. Grow your startup through viral spread, paid advertising or repeat customers
If you want to grow a profitable business, then this next part should be very exciting to you. Now we’ll talk about the ‘three growth engines’ that Ries says drive sustainable growth in every successful startup.
1. The first way to grow is through viral spread.
Viral spread is when your users introduce your product to new people one-by-one. This can include:
- Word of mouth. People love sharing cool things with friends. Facebook exploded in popularity through word of mouth, first on college campuses, then high schools and then everywhere else.
- A referral program. When Paypal launched, they gave $20 for referring a friend. This brought them a lot of users very quickly.
- Exposure as a side effect of using the product. This is when your customers use your product publicly, almost accidentally spreading the word. For example, the email service Hotmail wasn’t very popular… UNTIL they added a link to the bottom of every one of their user’s emails that said “Get your free email account at Hotmail.” So every time a Hotmail user sent an email, they were advertising Hotmail to their contacts. A year and a half later, Hotmail had jumped to 12 million users and Microsoft bought them for $400 million.
There’s actually a number called the viral coefficient that measures how viral or contagious a product is. For example, a viral coefficient of 2 means every user tells 2 of their friends about your product. If you are relying on viral spread for your growth, then you want this number as high as possible.
By the way, if you want to learn how to make your social media content more viral, then a VERY useful book is Contagious by Jonah Berger, for which we have a summary too. Jonah Berger breaks down 6 principles based on science that make people more likely to share your posts. Now back to The Lean Startup…
2. The second way to grow is through paid advertising.
Many businesses grow by paying for advertising which attracts new customers. We all know this.
The most important part of this is making sure that you’re paying LESS for each new customer than the PROFIT they’ll bring you. Does that make sense? You don’t want to spend $100 on advertising and then only make $50 in sales. In technical terms, business people would say that your LTV (customer lifetime value) must be lower than your CPA (cost per acquisition).
Eric Ries says IMVU grew because they figured out how to make money from a group of customers that were very cheap to advertise to online, including teenagers, broke people and those outside first world countries. Advertising to these groups was very cheap because other companies couldn’t make money from them.
As a side note, one of my favourite marketing teachers Dan Kennedy has a strategy that flips this around. He often says that “the best competitive advantage your business can have is being able to pay more than anyone else to acquire a customer” (paraphrased) That means if your business can make a high profit per customer, then you’ll be able to afford most types of advertising easily. Many businesses are always looking for the cheapest source of traffic, but this strategy totally flips that around.
3. The third way to grow is through repeat customers.
This is about holding onto your existing customers longer.
Paid subscriptions are ideal for this, because your customers pay automatically each month. So your revenue will continue to grow, as long as not too many people cancel the subscription and you’re getting some new customers.
Churn rate is the technical term for what percentage of your customers cancel every month in a subscription. The goal here should be to reduce your churn rate.
Adobe multiplied their gross profits by switching to a subscription business model. They went from $3 billion in 2013… to $9 billion in 2019! For a long time, they had always charged a one-time payment of hundreds of dollars for their software. Now they charge $10-50/month and they are far more successful. Other subscription services like Netflix, Disney+ and Spotify are also doing really well.
So those are the 3 growth engines! Pretty exciting, isn’t it?
Always remember that the growth engine you choose will determine what numbers you need to focus on in your business. These are your actionable metrics. For example, if you want to grow your business through viral spread, then get that viral coefficient number as high as possible.
Every successful startup has grown in one of three ways. First through viral spread like includes word of mouth or customer referrals. Second, through paid advertising, where you just have to make sure you making more money than what you’re spending for the advertising. Third, through customers buying from you repeatedly like with paid subscriptions. The growth engine you choose will determine what metrics you need to focus on optimizing.
7. Pivot to a new business strategy when your actionable metrics stop improving
What differentiates the success stories from the failures is that the successful entrepreneurs had the foresight, the ability, and the tools to discover which parts of their plans were working brilliantly and which were misguided, and adapt their strategies accordingly.
Startups often need to pivot.
Pivoting means keeping your team the same, and usually your product the same, but trying a different business strategy. In basketball, a pivot is when a player keeps one foot stuck to ground, but they rotate the rest of their body to face whichever direction they want. I think that’s a good way to visualize a pivot.
There are many ways a business can pivot, including:
- Finding a new group of customers for your existing product. For example, trying to sell your patient scheduling program to dentists instead of doctors.
- Sticking with your current group of customers and building a totally new product for them. This can happen when you learn a lot about a certain type of customers during the course of creating your first unsuccessful product.
- Focusing on one killer feature of your product and throwing out the rest. Basecamp.com is a great example of a software company that stopped working on several smaller apps so they could focus totally on their one best product.
- Using a different sales channel, like switching from selling to consumers to selling to businesses with an enterprise product.
One company that has a fantastic story of successfully pivoting is Nike. (This is a personal favourite of mine.) So Phil Knight started out being just the US distributor for a Japanese shoe company called Onitsuka Tiger. He sold their shoes in America for them and he was very successful at it. But one day the Japanese company said they were switching to a new distributor in America. That meant Phil had employees, stores and a lot of expenses, but no shoes. A big problem, to say the least.
In a rush, Phil started his own brand that would later be called Nike. He ordered several thousand pairs of custom shoes from an overseas factory. In just a few weeks, he pivoted from being a shoe distributor to selling his own brand of shoes. And we all know how this story turns out. If you want to listen to the whole inspirational business story behind Nike, then go read our summary of the book Shoe Dog by Phil Knight. I highly highly recommend that book to any aspiring entrepreneur.
So when should you pivot? Eric Ries says when your actionable metrics are not improving no matter what you try, then you probably need to pivot. Then, if your actionable metrics begin improving again after the pivot, that is a good sign you’ve made a successful pivot.
A pivot means you keep the same team and often the same product, but change your business strategy. For example, maybe you try to market your product to a new group of customers or you try a different sales channel. You must pivot when your actionable metrics are no longer improving despite your best efforts.
That’s all for this look at The Lean Startup. Lots of fascinating ideas in this book.
Next I recommend you look at our notes for The 22 Immutable Laws of Marketing. If you’re interested in creating a business strategy that makes you stand out from everyone else from the very beginning, then that book is invaluable.